Implementation of Deep Neural Networks in Facial Emotion Perception in Patients Suffering from Depressive Disorder: Promising Tool in the Diagnostic Process and Treatment Evaluation

Abstract

According to World Health Organization, depression is a common illness worldwide, with more than 300 million sufferers. This article describes relatively new research that is giving Deep Neural Networks (DNN) and Expert System-based hybrid solutions, skills of recognizing human affect and its intensity in standardized manner with more precision and objectivity than human eye. At present diagnostic process of depression relies mostly on diagnostic and statistical manual (DSM-5) and international statistical classification of mental disorders (ICD-10) alongside other standardized clinical measures conducted by clinicians. Implementation of DNN in recognition facial affect in depression appears a promising diagnostic tool, in conjunction with above mentions classifications and clinical measures, via improving early detection of depressive symptoms or facilitating evaluation of treatment efficacy in depressive disorder. The article particularly aims at automatic analysis of facial affect in depressed individuals, highlighting applications together with challenges to their implementation in medicine.